Dried blood spots in therapeutic drug monitoring and toxicology

KEYWORDS:

• Alternative sampling strategies
• dried blood spots
• microsampling
• therapeutic drug monitoring
• toxicology

1. Introduction

Dried blood spots (DBS) are generally obtained after a finger- or heel prick, by depositing a drop of capillary blood on a dedicated filter paper. While recently also other types of devices capable of generating dried blood samples (strictly taken not DBS) have entered the market [1], this Expert Opinion will refer to DBS as a prototype of dried blood samples, for reasons of simplicity. DBS have been used for highly diverse applications, starting from newborn screening, over therapeutic drug monitoring (TDM) throughout toxicology [2–9]. The growing interest in DBS sampling during the last decade is resulting from the many advantages coupled to this sampling technique. Since DBS are mostly obtained by a finger prick, sampling can be performed by the patient himself, eliminating the need for a phlebotomist. Furthermore, the small sample volume is beneficial when envisaging use in a pediatric or anemic population. The noncontagious nature of DBS and the fact that the dried matrix increases analyte stability poses fewer difficulties in terms of transport and storage. Besides, the sample preparation procedure is also straightforward and amenable to automation.

DBS also suffer from some challenges, with the hematocrit (hct) effect certainly being the most discussed issue [10]. In essence, because of a different viscosity, blood with lower respectively higher hct will spread more respectively less than control blood, typically leading to under-respectively overestimation of analyte concentrations when analyzing a DBS subpunch. Furthermore, capillary concentrations can be different from venous concentrations. Other limitations coupled to DBS sampling include the dependence on adequate sampling, risk of contamination, influence of the spotted volume, and spot inhomogeneity. This imposes a more complicated analytical and clinical validation procedure [4,6,10]. In addition, due the small sample size (typically 3–12 µl) associated with DBS, the amount of analytes and the number of analyses is limited. Therefore, sensitivity requirements may not always be met by available analytical instrumentation.

In recent years, many efforts have been made to facilitate correct dried blood sampling, to improve DBS bioanalysis and to cope with challenges related to DBS analysis. New sampling devices are entering the market, automated analysis leads to a higher efficiency, sample throughput and reproducibility and different strategies to cope with the hct effect have been put forward [10,11]. Innovations include whole-spot analysis of volumetrically applied DBS, development of alternative dried blood sampling devices, introduction of special types of filter paper, as well as cards for generating dried plasma spots, and setup of approaches for hct prediction of DBS, allowing the correction of hct-skewed results [11,12].

Here, we will focus on the use of DBS in the fields of toxicology and TDM.

2. Toxicology

DBS have been applied in a wide range of applications in toxicology, covering fields as toxicokinetics, epidemiology, environmental, and forensic toxicology [7–9,13]. Analytes measured in this context comprise (markers of) substances of abuse, environmental contaminants, and (trace) elements [7–9]. The potential of prompt and easy sampling and increased analyte stability offered by DBS sampling can be of key importance in several contexts, for example, roadside testing (driving under the influence of drugs (DUID) cases), workplace monitoring, drug-facilitated sexual assault cases, follow-up of drug and alcohol addicts, etc. [7–9,14].

The use of DBS in the context of DUID has only been evaluated to a limited extent. As compared to urine collection, DBS sampling is gender neutral and not accompanied by privacy or adulteration issues. In several countries, oral fluid has become the standard matrix for both screening and confirmation for roadside testing. However, also oral fluid sampling is subject to issues and interpreting oral fluid concentrations is not as straightforward as interpreting blood concentrations (which correlate best with intoxication). Moreover, DBS are considered noncontagious, which is advantageous for follow-up of several populations with a high prevalence of viral infections (e.g. HIV and hepatitis) [7,8].

The stabilizing effect of DBS – slowing down degradation and/or preventing de novo formation – has also proven to be beneficial in the field of toxicology. A nice example is the prevention of de novo formation of phosphatidylethanol, a marker indicative of alcohol use, which can be formed ex vivo in blood in the presence of ethanol [14]. In forensic toxicology, DBS may serve as a sample preparation strategy and offer the potential to store evidence of closed cases in a cost-effective way [7,8].

DBS offer the potential to study biomarkers of exposure, including DNA adducts and protein adducts [13]. Several protein adducts – primarily with albumin and hemoglobin – have already been quantified using DBS [13]. The features associated with DBS sampling and handling make DBS-based adduct monitoring particularly attractive for large-scale epidemiological studies, an application field that will likely continue to grow in future [13].

3. Therapeutic drug monitoring

TDM is another field with growing interest in the use of non- and minimally invasive alternative sampling strategies [4–6]. The potential for remote or home sampling, combined with the noncontagious character of DBS, makes it very attractive for TDM. Following sample collection at home, the patients can send the obtained DBS via regular mail to the analyzing laboratory, allowing lab results to be available before a patient visits a clinician for routine follow-up – or even rendering a routine follow-up consultation superfluous.

DBS-based TDM has been reported for various therapeutic drug classes, including anticonvulsants, antiretrovirals, immunosuppressants, antimalarials, antibiotics, analgetics, antidepressants, etc. [4–6,12]. Also (bio)markers such as HbA1c can be followed up via dried blood microsamples [15]. However, it needs to be mentioned that routine implementation in clinical laboratories has remained limited so far [12]. The small amount of sample mostly requires very selective and sensitive instrumentation (e.g. LC-MS/MS), although for selected analytes also other techniques, among which LC with UV and fluorescence detection and immunoassays have been used.

The development of fully automated DBS analyzers that can be directly coupled to standard LC-MS/MS configurations is essential for routine laboratories focusing on high-throughput assays. In these systems, solvents elute a fixed area from a DBS card, obviating the punching step, after which the extract is directed to an online sample clean-up and/or separated on an LC column, or even directly injected into the MS. By built-in accessories, such as, for example, barcode readers, these automated systems eliminate human errors associated with manual handling, allowing sample registration and traceability [12].

4. Expert opinion

In the context of TDM, we believe there is a bright future for DBS, for several reasons. First, there is an increased demand from patients to allow drug monitoring at home – DBS may perfectly serve this purpose. Second, there is an increased financial pressure on social security systems: DBS-based monitoring may be a cost-effective approach, not only allowing patient-tailored therapy, but also serving to check patient compliance. Indeed, with several medications costing thousands of dollars per month (e.g. oncology drugs), we have come beyond the point of having the right drug for the right patient: instead we wish to attain the right concentration of the right drug at the right time in the right patient. Third, automated processing of DBS can be made fully compatible with fully automated LC-MS/MS workstations that are about to enter clinical laboratories, allowing ‘from card to report, with no manual intervention’. These systems may be equipped with tools for automated recognition of correctly deposited DBS and for coping with variables like hct, for example, via the integration of reflectance spectroscopy for hct prediction [11]. This tailoring to current (and future) workflows of routine clinical laboratories, focusing on high-throughput, is essential for more widespread implementation [12].

In the context of toxicology, microsampling will continue to be used for animal and human studies, albeit priority may be given to liquid microsampling when expertise in dedicated sampling is present. Also in an acute (clinical) toxicology or TDM setting, liquid microsampling might be preferred over DBS sampling when feedback is urgent. Moreover, in those cases, sampling will typically be performed in a hospital setting, where it would actually not be logical to choose a DBS approach (needing to wait for a sample to dry). Yet, even in a hospital context, the sampling itself may offer opportunities or advantages, as exemplified by, for example, HbA1c monitoring of ‘wet’ capillary blood samples [15]. In general, the choice for DBS sampling and analysis needs to be well-balanced, taking into account the clinical question and the context in which sampling and analysis need to take place.

More toxicological applications where DBS may increasingly be used include screening approaches (if there are no time constraints, as stated above) as well as follow-up of markers of clinical or forensic interest. Examples include the follow-up of problematic drinkers (via direct alcohol markers) or drug users, where supervised sampling may be done, as well as follow-up of trace elements (e.g. in prosthesis patients) or of biomarkers of exposure in epidemiological surveys [7–9,13,14]. Although the sensitivity offered by the newest generation of analytical instrumentation is in many cases sufficient, for some applications an extra level may be warranted.

We believe the momentum has come for DBS to enter the main stage of TDM and toxicology: the proof of principle has been convincingly established for many applications and it is now time for a more widespread implementation. Of course, strict quality criteria need to be taken into account if DBS sampling and analysis is to be used for clinical or legal decision-making. Initiatives to set up proficiency testing programs, as well as widely accepted best practice guidelines on analytical and clinical validation for DBS-based methodologies are essential for further integration of DBS-based methods into routine practice [6,10,12]. Last, it should be realized that DBS sampling and analysis is not suitable for every analyte in every context and will never replace traditional sampling: DBS sampling and analysis should rather be looked at as an additional tool in the analytical toolbox [12]. If properly performed, it has the ability to provide high-quality results where adequate information cannot be (conveniently) obtained using traditional procedures (e.g. sample collection in remote or resource-limited areas and at patient’s homes).

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. Peer reviewers on this manuscript have no relevant financial or other relationships to disclose

Additional information

Funding

SV holds a PhD fellowship funded by the Special Research Fund (BOF) [01D42414].